Advanced Fracturing Fluids

Improve Well Economics

The oil and gas industry has witnessed a revolution in fluids technology

for hydraulic fracturing. Starting in the mid 1980s, focused research led

to major improvements in the performance of well stimulation fluids.

Today, new additives and fluids are extending these capabilities and

providing innovative solutions to nagging problems. The results are

more efficient and cost-effective treatments for enhancing well produc-

Kevin Armstrong Hydraulic fracturing is one of the oil and gas

Roger Card industry’s most complex operations. This
Reinaldo Navarrete technique has been applied worldwide to
Erik Nelson increase well productivity for nearly 50
Ken Nimerick years.1 Fluids are pumped into a well at
Michael Samuelson pressures and flow rates high enough to split
Tulsa, Oklahoma, USA the rock and create two opposing cracks
extending up to 1000 ft [305 m] or more
Jim Collins from either side of the borehole (right ). Sand
Calgary, Alberta, Canada or ceramic particulates, called proppant, are
carried by the fluid to pack the fracture,
Gilbert Dumont keeping it open once pumping stops and
Michael Priaro (consultant) pressures decline.
Neal Wasylycia What defines a successful fracture? It is
Petro-Canada Resources one that:
Calgary, Alberta, Canada • is created reliably and cost-effectively
• provides maximum productivity enhance-
Gary Slusher ment
Enron Oil & Gas • is conductive and stable over time.
Corpus Christi, Texas, USA This article describes today’s fracturing
operations and the pivotal role played by
the fracturing fluid. Then, it highlights four
new fluid technologies that are improving
fracture success and well economics.

For help in preparation of this article, thanks to Vic
Joyce, Ken Nolte and Jon Mitchell, Dowell, Tulsa,
Oklahoma, USA; Terry Greene, Dowell, Montrouge,
France; Richard Marcinew, Dowell, Calgary, Alberta,
Canada; and Mike Morris, Dowell, Corpus Christi,
Texas, USA.
In this article, HIGHSHEAR, CleanFLOW and PropNET
are marks of Schlumberger.
1. For more on hydraulic fracturing:
Waters AB: “Hydraulic Fracturing—What Is It?,” Jour-
nal of Petroleum Technology 33 (August 1981): 1416.
Veatch RW Jr, Moschovidis ZA and Fast CR: “An
Overview of Hydraulic Fracturing,” in Gidley JL,
Holditch SA, Nierode DE and Veatch RW Jr (eds):
Recent Advances in Hydraulic Fracturing, Monograph
12. Richardson, Texas, USA: Society of Petroleum
Engineers (1989): 1-38.

34 Oilfield Review
 Pump trucks
Continuous fluid mixer Proppant blender

nFracturing operation. Modern treatments rely on a variety of process-controlled mixing, blending and pumping equipment
and computer monitoring and recording systems, which permit real-time decision making.

Autumn 1995 35
Unfractured, radial flow Low-permeability, linear flow

High-permeability, linear flow

n Flow regimes. Radial flow profile for an unfractured well
(top left). Linear flow into and along the fracture replaces radial
flow with a more conductive path for hydrocarbons (top right). In
high-permeability formations, short fractures are used to reach
beyond the area of matrix damage near the wellbore (bottom).

The Rock, the Mechanics and the Fluid
Historically, fracturing has been applied pri-
marily to low-permeability—0.1 to 10 md—
formations with the goal of producing nar-
row, conductive flow paths that penetrate
deep into the reservoir. These less restrictive
linear conduits replace radial flow regimes
and yield a several-fold production increase
(above ). For large-scale treatments, as many
as 40 pieces of specialized equipment, with
a crew of 50 or more, are required to mix,
blend and pump the fluid at more than 50
barrels per minute (bbl/min) [8 m 3/min].
Pumping may last eight hours with
1,000,000 gal of fluid and 2,000,000 to
4,000,000 lbm of proppant placed in the
fracture (next page, left ) .
Until recently, treatments were performed
almost exclusively on poor producing wells
(often to make them economically viable).
In the early 1990s, industry focus shifted to
good producers and wells with potential for
greater financial return. This, in turn, meant
an increased emphasis on stimulating high-
permeability formations.
The major constraint on production from
such reservoirs is formation damage, fre-
quently remedied by matrix acidizing treat-
ments. But acidizing has limitations, and
fracturing has found an important niche.
The objective in highly permeable forma-
tions is to create short, wide fractures to
reach beyond the damage. This is often
accomplished by having the proppant
bridge, or screen out, at the end, or tip, of
the fracture early in the treatment. This “tip
screenout” technique is the opposite of what
is desired in low-permeability formations
36
where the tip is ideally the last area to be
packed.
Why the different approach? The answer is
found in the relationship between fracture
length and the permeability contrast
between the fracture and the formation.
Where the contrast is large, as for low-per-
meability reservoirs, longer fractures provide
proportionally greater productivity. Where
the contrast is small, as in high-permeability
formations, greater fracture length provides
minimal improvement. Fracture conductivity
is, however, directly related to fracture
width. Using short—about 100-ft [30-m]—
and wide fractures can prove beneficial.2
High-permeability formation treatments
are on a far reduced scale. Only a few
pieces of blending and pumping equipment
are required, and pumping times are typi-
cally less than one hour, and often only 15
minutes. Fluid is pumped at 15 to 20
bbl/min [2.4 to 3.2 m 3/min] with a total
volume of 10,000 to 20,000 gal [37.9 to
75.7 m3] and total proppant weight of about
100,000 lbm [45,000 kg] (next page, left ).
This technique has been successful in the
North Sea, Middle East, Indonesia, Canada
and Alaska, USA.
While fracturing treatments vary widely in
scale, each requires the successful integra-
tion of many disciplines and technologies,
regardless of reservoir type. Rock mechanics
experiments on cores, specialized injection
testing and well logs provide data on forma-
tion properties. Sophisticated computer soft-
ware uses these data, along with fluid and
well parameters, to simulate fracture initia-
tion and propagation. These results and eco-
nomic criteria define the optimum treatment
design. Process-controlled mixing, blending
and high-pressure pumping units execute
the treatment. Monitoring and recording
devices ensure fluid quality and provide
permanent logs of job results. Engineers
tracking the progress of the treatment use
graphic displays that plot actual pumping
parameters against design values to facilitate
real-time decision making. Production simu-
lators compare treatment results with expec-
tations, providing valuable feedback for
design of the next job.
At the heart of this complex process is the
fracturing fluid.3 The fluid, usually water-
based, is thickened with high molecular
weight polymers, such as guar or hydrox-
ypropyl guar. It must be chemically stable
and sufficiently viscous to suspend the prop-
pant while it is sheared and heated in sur-
face equipment, well tubulars, perforations
and the fracture. Otherwise, premature set-
tling of the proppant occurs, jeopardizing
the treatment. A suite of specially designed
chemical additives imparts important prop-
erties to the fluid. Crosslinkers join polymer
chains for greater thickening, fluid-loss
agents reduce the rate of filtration into the
formation and breakers act to degrade the
polymer for removal before the well is
placed on production.
The fracture is created by pumping a
series of fluid and proppant stages. The first
stage, or pad, initiates and propagates the
fracture but does not contain proppant. Sub-
sequent stages include proppant in increas-
ing concentrations to extend the fracture
and ensure its adequate packing.
Fracturing fluid technology has also devel-
oped in stages. Early work focused on iden-
tifying which polymers worked best and
what concentrations gave adequate prop-
pant transport. Then, research on additives
to fine-tune fluid properties hit high gear.
Oilfield Review
 Much was learned, but what finally
40 vehicles
emerged was a huge array of complicated
fluids—difficult to prepare and pump—and
an amazing assortment of single-use addi-
tives (most had to be custom manufactured)
that required expensive material inventories.
In the past ten years, a more productive
research direction has emerged. Oil compa-
1,000,000 gal nies, service companies and polymer manu-
facturers have concentrated on the basic
physical and chemical mechanisms underly-
ing the behavior of fracturing fluids in an
attempt to find improved approaches to fluid
design and use. This initiative has led to
major advances, including higher-performing
polymers, simpler fluids, multifunctional
additives and continuous, instead of batch,
mixing. These developments have had a sig-
3 million lbm
nificant, beneficial impact on the industry.
Recent innovations are extending the state
Propp Propp Propp
of the art in four areas:
Propp Propp Propp • controlling fluid loss to increase
fluid efficiency
Propp Propp Propp • extending breaker technology to
improve fracture conductivity
Propp Propp Propp
• reducing polymer concentration to
improve fracture conductivity
Propp Propp Propp
• eliminating proppant flowback to
stabilize fractures.
Propp Propp Propp
Each provides new opportunities for
Typically, particulates or other fluid addi-
tives are used to reduce leakoff by forming a
filter cake—termed an external cake—on
the surface of the fracture face. Acting
together with the polymer chains, the fluid-
loss material blocks the pore throats, effec-
tively preventing invasion into the rock
matrix (below ) .
This approach has been applied suc-
cessfully for decades to low-permeability
Flow
External filter cake
Bridging zone
Invaded zone
Uncontaminated
formation
nZones of leakoff in a fractured formation.
In low-permeability reservoirs, the
invaded zone may be small or nonexis-
tent, since polymer molecules are too
large to enter the matrix. In these cases,
an external filter cake controls fluid loss.
In high-permeability reservoirs, signifi-
cant matrix damage can occur because
of particle penetration.

Propp Propp Propp

improving well economics, as described in
the remainder of this article. 2. Hanna B, Ayoub J and Cooper B: “Rewriting the Rules
for High-Permeability Stimulation,” Oilfield Review 4,
Propp Propp Propp
no. 4 (October 1992): 18-23.
Controlling Fluid Loss 3. For background on fracturing fluids: Constien VG:
Propp Propp Propp A portion of the fluid pumped during a frac- “Fracturing Fluid and Proppant Characterization,” in
Economides MJ and Nolte KG (eds): Reservoir Stimu-
turing treatment filters into the surrounding lation, 2nd ed. Englewood Cliffs, New Jersey, USA:
Propp Propp Propp
permeable rock matrix. 4 This process, Prentice Hall (1989): 5-1–5-23.
referred to as fluid leakoff or fluid loss, 4. Penny GS and Conway MW: “Fluid Leakoff,” in
Propp Propp Propp
occurs at the fracture face. The volume of Gidley JL, Holditch SA, Nierode DE and Veatch RW Jr
(eds): Recent Advances in Hydraulic Fracturing,
fluid lost does not contribute to extending or Monograph 12. Richardson, Texas, USA: Society of
Propp Propp Propp
widening the fracture. Fluid efficiency is one Petroleum Engineers (1989): 147-176.
parameter describing the fluid’s ability to 5. Fluid efficiency, E, is defined as
Propp Propp Propp
create the fracture.5 As leakoff increases, VF VF
E= =
Propp Propp Propp
efficiency decreases. Excessive fluid loss can V P V F +V L
jeopardize the treatment, increase pumping
in which VF is the fluid volume remaining in the
Propp Propp Propp costs and decrease post-treatment well per- fracture, VP is the total volume pumped and VL is the
formance. volume of fluid that has leaked into the formation.
Propp Propp Propp

Propp Propp Propp 4 vehicles

Propp Propp Propp

20,000 gal 100,000 lbm

Propp Propp Propp Propp

Propp Propp Propp Propp

Equipment Fluid Proppant Equipment Fluid Proppant

Low-Permeability Treatment High-Permeability Treatment

nScale of fracturing treatments. Equipment, fluid and proppant requirements to create

long fractures in low-permeability formations (left) can be 10 to 50 times more than
those needed for short fractures in high-permeability reservoirs (right).

Autumn 1995 37
 Measuring Dynamic Fluid Loss
in the Laboratory

Side View Fluid inlet End View

Oil-filled annulus
Fluid To differential
slot pressure
transducers Sleeve

Leakoff Core
fluid

Core
holder

Core Sleeve Oil-filled annulus

Fluid outlet

■Dynamic fluid-loss cell and core holder. A specially designed core holder simulates a slot-flow geometry, the one most representative of the actual fracturing
process. Pressure and leakoff measurements provide the data necessary to evaluate the depth of invasion and the impact of various additives on fluid loss.

Dynamic fluid-loss measurements were made in ■Dynamic fluid-loss

the Dowell laboratory in Tulsa, Oklahoma, USA test apparatus.
using special fluid-loss cells with a slot-flow
geometry and a porous test surface on one of the
slot walls (above). Cylindrical cores of the same
type and dimensions as for static tests are used
to allow direct comparison with static fluid-loss
results. Cells are constructed from stainless steel
for operation to 3500 psi and 350°F [177°C]. The
inlet design ensures fully developed flow over the
test section. A backpressure regulator and a heat
exchanger, which cools the filtrate, prevent evap-
oration of the filtrate during operations above the
ambient boiling point of the fluid. As many as
three cells can be used simultaneously for testing
cores of differing permeabilities.
A special fluid-loss simulator was designed to
prepare the fluids under dynamic conditions, sub-
ject them to shear and temperature histories and
then measure fluid loss (right).

The two components of the apparatus are a
shear history simulator and a fracture simulator.
The first uses a static mixer and 800 ft [244 m] of
small-diameter tubing to simulate preparation
and shearing of the fluid in the well tubulars.
The second subjects the fluid to shear and temper-
ature conditions of the fracture, using two large,
38
floating-piston accumulators and coils of tubing
immersed in a temperature-controlled bath.
A computer-controlled valving arrangement
ensures that the fluid always travels in the same
direction across the core face.
Oilfield Review

(< 0.1 md) formations in which polymer
and particulate sizes exceed those of the
pore throats. In high-permeability reservoirs,
however, fluid constituents may penetrate
into the matrix, forming a damaging internal
filter cake. This behavior has prompted
mechanistic studies to determine the impact
on fracturing treatment performance.6
Classic fluid-loss theory assumes a two-
stage, static—or nonflowing—process.7 As
the fracture propagates and fresh formation
surfaces are exposed, an initial loss of fluid,
called spurt, occurs until an external filter
cake is deposited. Once spurt ceases, pres-
sure drop through the filter cake controls
further leakoff. For years, researchers have
developed fluid-loss control additives under
nonflowing conditions based on this theory.
The conventional assumptions, however,
neglect critical factors found under actual
dynamic—or flowing—conditions present
during fracturing, including the effects of
shear stress on both external and internal fil-
ter cakes and how fluid-loss additives move
toward the fracture face. In high-permeabil-
ity formations, with an internal filter cake
present, most of the resistance to leakoff
occurs inside the rock, leaving the external
cake subject to erosion by the fluid.
Analysis of fluid loss under dynamic condi-
tions relates external cake thickness to the
yield stress of the cake at the fluid interface
and the shear stress exerted on the cake by
the fluid.8 These, in turn, depend on the
physical properties of the cake and the rheo-
logical properties of, and shear rate induced
in, the fluid.9 Whether an external filter cake
forms, grows, remains stable or erodes
depends on the way these parameters vary
and interact over time and spatial orientation.
Similarly, the effectiveness of additives to
control fluid loss depends on two factors:
their ability to reach the fracture face quickly
and their ability to remain there. The former
is governed by the drag force exerted on the
particles and the latter by the shear force
exerted on them (right ) . The larger the ratio
of drag to shear, the greater the chance that
the particles will remain on the surface. A
greater leakoff flux to the wall, smaller parti-
cle dimensions and a lower shear rate favor
sticking. Promoting higher leakoff for better
additive placement seems directly at odds
with controlling fluid loss! However, in prac-
tice, higher initial leakoff can yield greater
overall fluid efficiency.
To confirm the controlling mechanisms,
dynamic fluid-loss tests were conducted
Autumn 1995
350
300
Calculated
Laboratory approximation
250
Fracture height = 300 ft
Shear rate, sec -1
Fracture length = 540 ft
200
Pump rate = 40 bbl/min
Pump time = 145 min
150
100
50
0 20 40 60 80 100 120 140 160
Time, min
■ Shear rate history of a rock segment. As the tip of the fracture passes a particular
location, in this case a point 50 ft [15 m] from the wellbore, the maximum shear rate
occurs. As the treatment progresses and the fracture widens at this location, the shear
rate falls off rapidly initially and then more slowly later. The laboratory approximation
and calculated curve show good agreement.
using a slot-flow geometry, determined to be
6. Navarrete RC, Cawiezel KE and Constien VG:
the simplest representation of what occurs in “Dynamic Fluid Loss in Hydraulic Fracturing Under
a fracture. To completely describe the pro- Realistic Shear Conditions in High-Permeability
Rocks,” paper SPE 28529, presented at the 69th SPE
cess, computer-controlled equipment was Annual Technical Conference and Exhibition, New
constructed to prepare and test fluids under Orleans, Louisiana, USA, September 25-28, 1994.
dynamic conditions, subjecting them to the 7. Howard GC and Fast CR: Hydraulic Fracturing,
temperature and shear histories found in a Monograph 2. Richardson, Texas, USA: Society of
Petroleum Engineers, 1989.
fracture (see “Measuring Dynamic Fluid Loss 8. Prud’homme RK and Wang JK: “Filter-Cake Forma-
in the Laboratory,” previous page ). Cores of tion of Fracturing Fluids,” paper SPE 25207, pre-
various lengths were used in the tests to sim- sented at the SPE International Symposium on Oil-
field Chemistry, New Orleans, Louisiana, USA,
ulate a fracture segment at a fixed distance March 2-5, 1993.
from the wellbore. 10 As the fracture tip 9. Shear rate is proportional to shear stress, the exact
passes a specific point, spurt occurs and the relationship depending on the rheological model
used for the fluid.
shear rate reaches a maximum ( above ).
10. Navarrete RC and Mitchell JP: “Fluid-Loss Control
Then, as the fracture widens, the shear stress for High-Permeability Rocks in Hydraulic Fracturing
decreases. In the test apparatus, this is simu- Under Realistic Shear Conditions,” paper SPE
lated by decreasing the flow rate with time. 29504, presented at the SPE Production Operations
Symposium, Oklahoma City, Oklahoma, USA, April
2-4, 1995.
vx
y
Fx
Fy
■ Forces acting on a particle. A particle,
such as a fluid-loss additive, flowing
inside a fracture is subjected to shear, Fx
and drag, FY , where vx is the velocity in
the flow direction as a function of the dis-
tance, y, from the fracture face. The ratio
of shear to drag is directly proportional to
the leakoff rate and inversely propor-
tional to the particle size and the shear
rate at the wall. High initial leakoff and
optimized particles can help ensure that
fluid-loss additives reach and remain on
the fracture face.
39
 Pressure sensors along the core monitor the
6 progress of the polymer front.
Temperature = 150°F
Laboratory tests show that, for compara-
Pressure drop = 1000 psi
ble fluids and rocks with permeabilities of
5 up to 50 md, fluid loss is greater under
Dynamic, 190 sec–1 shear rate
dynamic conditions than static conditions
(left ). Further, examining the impact of shear
4 stress and permeability on the magnitude of
Fluid-loss vol, ml fluid loss and the effectiveness of leakoff-
control additives in high-permeability for-
Dynamic, 40 sec–1 shear rate
3 mations led to five key conclusions.11
First, high shear rates can prevent the for-
mation of an external filter cake and result
2 Static in higher than expected spurt (below, left ).
Second, an internal filter cake controls fluid
loss, especially near the fracture tip. Third,
1 the effectiveness of fluid-loss additives
increases with formation permeability and
decreases with shear rate and fluid viscosity.
0 Fourth, reducing fluid loss means reducing
2 4 6 8 10 12 14 16 18
Time, min spurt, particularly under high shear condi-
tions and in high-permeability formations.
nDynamic versus static fluid-loss test results. For this crosslinked guar system, dynamic Finally, at high shear rates with no external
fluid-loss values are higher than static measurements. As shear rate increases, leakoff filter cake, efficient spurt control must be
increases. Here, rock permeability is 0.5 md. achieved by plugging the pore throats at the
surface of the rock.
16 350 The effect of shear depends on the type of
fluid and the formation permeability. Typi-
14 cally, above a threshold shear level, no
300
external filter cake is formed. The magni-
12
250
tude of fluid loss is dependent on the type of
62 md polymer and whether it is crosslinked. If the
10

Shear rate, sec-1

permeability is high enough and the fluid
Fluid-loss vol, ml

10 md
200
structure degrades with shear, polymer may
8
be able to penetrate the rock matrix.
150 Dynamic tests revealed that commonly
6
used additives were less effective in control-
100 ling fluid loss than static tests had previously
4
Shear rate indicated. Also, a direct link between fluid
50
efficiency and shear rate was demonstrated.
2
The higher the fraction of fluid lost under
high shear early in the treatment, the higher
0
0 the total leakoff volume and the lower the
20 40 60 80 100 120 140
efficiency. Spurt has a dominant effect on
Time, min
efficiency and the volume of fluid pumped,
nEffect of shear rate on dynamic fluid loss. Fluid loss increases with increasing perme- particularly for the short pumping times
ability, illustrating the importance of controlling leakoff in high-permeability forma- encountered in fracturing high-permeability
tions. Maintaining high fluid efficiency is critical to creating a cost-effective fracture formations. If spurt is not controlled quickly
with minimal formation damage. and effectively during the high-shear period,
fluid efficiencies drop dramatically.
These observations prompted researchers
to develop a superior additive system that:
• controls spurt under high shear rates in
high-permeability formations
• minimizes the influence of permeability
on leakoff12
• limits the invasion of polymer into
the matrix
• reduces the overall amount of polymer
pumped into the fracture.

40 Oilfield Review

Evaluation of several additive types led to
a combination of particulate materials, the
HIGHSHEAR system, that achieves the
above goals. One agent moves rapidly to
the fracture face during the early stage of
fluid loss when leakoff flux is high. This type
of particle seals a major portion of the
exposed surface quickly and adheres
securely to the surface, resisting high shear
forces. The second material plugs the
remaining gaps in the developing filter cake
as the shear rate drops, significantly reduc-
ing filtrate losses and sealing the surface so
that polymer particles cannot penetrate the
matrix. Laboratory tests show a 25 to 75%
reduction in spurt compared to the best
available products today (right ). Fluid effi-
ciency improves, meaning less fluid to
pump, less fluid to break and easier
cleanup. Since this process is aided by hav-
ing a less viscous fluid—one that promotes
high initial leakoff—there is a further oppor-
tunity for cost-savings and improved
cleanup by reducing the polymer concen-
tration in the fluid (page 46 ).
One concern with fluid-loss additives has
been the possibility of their presence reduc-
ing proppant-pack conductivity. A key ques-
tion is: Does the reduction in matrix dam-
age brought about by additives outweigh the
potential damage they may inflict within the
fracture itself? As is well-known, the amount
of conductivity damage is highly dependent
on the type of fluid used.13 Flowback tests
with different fluids and leakoff control
additives confirm that particulate fluid-loss
additives do limit matrix damage by mini-
mizing fluid invasion. Conductivity tests
show that these agents are actually less
damaging to the proppant pack than previ-
ously thought. In most fracturing fluids, their
impact on conductivity is minimal when
used in low concentrations.14
With these encouraging laboratory results,
the next step was to test whether the HIGH-
SHEAR system could reduce treatment costs
and improve well productivity better than
conventional products. Extensive field trials
in Canada show an overall 15 to 20% cost
savings (see “Canadian Treatments Demon-
strate Cost Savings and Productivity Gains,”
next page ). Productivity improvements aver-
aged 460% compared to 260% on offset
wells without the additive. Additional well
tests are planned in areas such as the US
Gulf Coast where a variety of fluids are used
and larger-scale, high-permeability fractur-
ing treatments are performed.
Autumn 1995
16
12
Fluid-loss vol, ml
8
4
0
10 20
Time, min
nEffect of additives on fluid loss. The HIGHSHEAR additive system provides substantial
improvement in controlling leakoff compared to the best materials available today.
Understanding and Improving Fracture
Conductivity
Simply creating a fracture does not guaran-
tee better well performance. The fracture
must provide a conductive flow path for for-
mation fluids. For decades, poor treatment
results were blamed on insufficient fracture
length or inadequate proppant transport and
placement. Studies in the late 1980s
revealed that substantial fracture damage,
and resulting impairment to flow, could be
caused by polymer residue blocking the
pore spaces between proppant particles.15
Once the proppant pack is placed and
pumping stops, fluid filtrate leaks off into the
rock matrix and the pressure declines. The
fluid remaining in the fracture must then be
flowed back to allow production of hydro-
carbons. Termed cleanup, this process is crit-
ical to the success of the treatment. One way
to aid cleanup is to ensure that the polymer
residue has been reduced to a minimum. A
second is to use as little polymer as possible
to start with. The next two sections look at
recent developments in both areas.
Extending Breaker Technology
At the end of a treatment, the fluid left in the
fracture has been partially dehydrated due to
filtrate loss. The effective polymer concentra-
tion can be an order of magnitude higher
than that originally pumped, reaching 300
to 600 lbm/1000 gal [36 to 72 g/cm3]. If the
polymer stays intact, an ultrahigh viscosity,
gelled mass results that blocks the pore
space and cannot easily be flowed back into
the well.
To prevent this, the polymer is attacked by
fluid breakers—oxidizers or enzymes that
sever the polymer chain at its weakest
points, degrading it into smaller, more
mobile fragments. This reduces the viscosity
No fluid-loss additive
25 lbm/1000 gal best available
15 lbm/1000 gal HIGHSHEAR
30 40 50
of the residual fluid, thereby allowing more
efficient cleanup. Breakers are used at reser-
voir temperatures below about 325°F
[163°C].16 If breaking is insufficient, con-
centrated polymer remains in the proppant
pack, reducing the conductivity and treat-
ment effectiveness.
Historically, active chemical breakers
were dissolved in the fluid during surface
mixing. As a result, the fluid was being
attacked even as it was being pumped. Care
had to be taken to use a sufficiently low
breaker concentration. Otherwise, viscosity
would decrease too quickly, and proppant
would settle. With this low breaker concen-
tration, only partial degradation of the poly-
mer occurred. The result was impaired frac-
ture conductivity.
Research during the past decade has
found ways to increase breaking efficiency.
11. Navarrete et al, reference 6.
12. In high-permeability formations, permeability can
vary widely. Using an average value in job design
calculations can present execution problems during
the treatment. An additive that dampens the effect of
permeability variation can be beneficial.
13. Brannon HD and Pulsinelli RJ: “Evaluation of the
Breaker Concentrations Required to Improve the Per-
meability of Proppant Packs Damaged by Hydraulic
Fracturing Fluids,” paper SPE 19402, presented at the
SPE Formation Damage Control Symposium,
Lafayette, Louisiana, USA, February 22-23, 1990.
14. In instances where the impact is significant, the dam-
age can be countered by lowering the polymer con-
centration since fluids of lower viscosity can be used
during the treatment.
15. Hawkins GW: “Laboratory Study of Proppant-Pack
Permeability Reduction Caused by Fracturing Fluids
Concentrated During Closure,” paper SPE 18261,
presented at the 63rd SPE Annual Technical Confer-
ence and Exhibition, Houston, Texas, USA, October
2-5, 1988.
16. Above this level, the fluid breaks with time due to the
thermal degradation of the guar molecules.
41
 Canadian Treatments Demonstrate
Cost Savings and Productivity Gains

More than 19 operational trials of the HIGHSHEAR

fluid-loss control additive have been conducted
in Alberta, Canada in clean sandstone reservoirs
(right). Here, permeabilities vary from 50 to
150 md and porosities are 20 to 24%. Well depths
typically range from 2300 to 2500 ft [700 to
760 m] with bottomhole temperatures of 85 to
95°F [30 to 35°C].
Earlier wells fractured in the field in 1993, as
part of a refracturing program, yielded proppant
placement efficiencies as low as 70%. Because of
the relatively high permeability, its variability ■Field trial of fluid-loss additive and low-guar fluid in Alberta, Canada.
within the reservoir and high-concentration prop-
pant schedules, an error of 10% in the fluid-loss 20
■Comparison of pro-
design calculation could lead to 30% of the prop-
18 duction rates before
pant not being placed. In addition to verification Rate before refrac
and after treatment.
of laboratory test results with the additive, objec- 16 Rate after refrac Five wells were refrac-
tured in 1993 using
Production rate, m3/day

tives for a new refracturing campaign were: 14

conventional fluids
• reduced treatment cost
12 technology, resulting
• improved job design and execution in an average 260%
• decreased volume of polymer pumped 10
improvement in pro-
• reduced well cleanup times 8 duction rate (top).
Nineteen wells refrac-
• increased production rates.
6 tured in 1995 using the
To achieve these goals, the strategy was to HIGHSHEAR fluid-loss
4
replace the 30 lbm/1000 gal borate-crosslinked additive and the low-
fracturing fluid normally used with a reduced- 2 guar fluid showed an
polymer, 22 lbm/1000 gal system (page 46) to average 460% improve-
0 ment (bottom).
promote higher fracture conductivity, to speed Well Well Well Well Well
1 2 3 4 5
cleanup (due to less damage from residual
polymer) and to lower wellsite costs. With its
inherently lower viscosity, the new fluid meant 30
higher leakoff. To counteract this, the HIGHSHEAR
additive was used to limit spurt, seal the Rate before refrac
25
fracture face and smooth out the effects of Rate after refrac
Production rate, m3/day

permeability variations.
20
Pad volumes of 2000 to 2500 gal [7.6 to 9.5m3]
were used with total slurry volumes of 4000 to
15
6000 gal [15.1 to 22.7 m3] in four stages. Typical
rates were 15.7 bbl/min [2.5 m3/min]. Maximum
sand concentrations ranged from 16.5 to more 10

than 18.5 ppa.

The field trials were highly successful. Test 5
wells showed consistent proppant placements of
95 to 100%. Over time, as experience with the 0
A B C D E F G H I J K L M N O P Q R S
system grew, pad volumes were maintained and
Wells

42 Oilfield Review
 Fractional retained permeability
By encapsulating breakers so they do not 1

interact with the fluid until released by rup- Theoretical model

0.8
ture of the protective coating with time or 1 Layer
stress, higher concentrations can be used, 0.6 2 Layer, test 1
degrading more of the polymer. This has 2 Layer, test 2
been a major innovation for improving 0.4

cleanup and proppant-pack conductivity

0.2
and is a common industry practice today.17
higher sand concentrations were achieved as a The mechanisms of proppant-pack dam-
0
result of increasing placement success. Less age have been the subject of much study 0.2 0.4 0.6 0.8 1

fluid pumped meant less fluid to break. Cleanup and much controversy. Recent findings have Fractional reduction in porosity

times were slashed from several days to one or

led to a more consistent and reproducible
■ Retained permeability as a function of
approach to testing proppant-pack damage. porosity. The theoretical correlation relat-
two days. Post-treatment production increased
By understanding the controlling processes, ing the two properties of the proppant
from 260% in the 1993 campaign to 460% (previ- scientists have been able to develop new- pack has been confirmed in laboratory
ous page, middle). Overall treatment costs were generation additives to improve fluid testing. Small decreases in porosity can
lead to significant reductions in retained
15 to 20% lower. cleanup and fracture conductivity. permeability, resulting in decreased
The changeover to the lower polymer loading A keystone of the investigation was devel- hydrocarbon production.
oping a relationship between retained per-
was successful due to the fluid-loss additive.
meability and porosity reduction in prop- fracture face lead to large reductions in
In separate trials without the material present, pant packs. A graph of this correlation retained permeability (below ).
proppant placement efficiency was erratic. shows that, for example, a 10% change in Another part of the study involved identi-
According to Gilbert Dumont, production engi- porosity (from 30 to 27%) can reduce per- fying the parameters and test procedures
neer for Petro-CanadaResources, “Fluids are meability by 35%. Laboratory tests verify the with the most significant effects on mea-
coming back crystal-clear with this system,
theoretical relationship (above, right ). Small, sured conductivity. Tests were performed at
random blockages by residual polymer or a the Dowell laboratory in Tulsa, Oklahoma,
something we haven’t seen before. The wells are
thin, concentrated polymer filter cake at the USA and two independent laboratories. The
cleaning up quicker, in one to two days rather
than three to ten days. This indicates we are get-
75% by filter cake
ting faster and more complete breaking of the 50%
75% retained permeability
25%
system. We believe our conductivities are higher
■ Pore blockages.
as supported by increased well productivity.”
Retained perme-
Michael Priaro, engineering consultant for ability can be
Petro-Canada Resources adds, “Our success rate reduced by an
external filter
is greater because of more consistent placement. cake (upper left).
We’ve been able to obtain higher proppant con- The larger the filter
cake, the lower
centrations with cleaner fluid. The cleanup has the permeability.
been better and quicker and well productivity Retained perme-
ability can also be
after stimulation is definitely higher, averaging reduced by random,
4.6-fold improvement, with some wells showing undegraded poly-
mer fragments fill-
almost 7-fold improvement. Ten-fold production
ing the void spaces
improvements may be possible with further between proppant
50% retained permeability 25% retained permeability
advances in treatment design.” particles. As the
residue increases,
From Neal Wasylycia, production engineer for the permeability
Petro-Canada Resources, “Comparisons of decreases (upper
right, lower left, lower
results are difficult with slant, deviated and right). Even small
directional wells with varied pays being frac- increases in residue
can have a dra-
tured. We definitely know that our results on ver- matic effect and
tical wells are dramatically better, and 100% in both cases,
the permeability
fracture placement can be repeated with confi-
decrease can dra-
dence. Our overall refracturing costs are reduced matically impact
due to lower fluid cost, quicker cleanup and production rates.

reduced fluid disposal cost. The successful

placement of fractures in deviated and slant wells
represents our next challenge, as this is where
our next large refracturing opportunity exists.”
Autumn 1995
17. Gulbis J, King MT, Hawkins GW and Brannon HD:
“Encapsulated Breaker for Aqueous Polymeric Flu-
ids,” paper SPE 19433, presented at the SPE Forma-
tion Damage Control Symposium, Lafayette,
Louisiana, USA, February 22-23, 1990.
43
results confirm that polymer concentration,
for example, is critical. Its influence is non-
linear (below ).
The relationships of other parameters to
test results and reproducibility were also
quantified, leading to a refined testing pro-
cedure. Testing by these laboratories using
the new guidelines demonstrates that reli-
able results can be achieved, provided test
conditions are closely controlled. Testing
within the same laboratory is expected to
vary by 15 to 20% while lab-to-lab varia-
tions have a 25 to 30% relative deviation
( right ) . Before uniform procedures, test
results often deviated by 100% or more.
100
Borate crosslinked guar
Retained permeability, %
Temperature = 160°F
80
60
40
20
0
0 100 200
Polymer concentration, lbm/1000 gal
nEffect of polymer concentration on
retained permeability. As the residual
polymer concentration remaining in the
proppant pack increases, the permeability
decreases, emphasizing the need for com-
plete polymer degradation and removal.
nEnhanced proppant-pack cleanup. Use of a breaker alone (top)
results in polymer fragments that may be difficult to remove
from the fracture. The CleanFLOW additive system works with
the breaker (bottom) to prevent fragment coagulation, facilitate
cleanup and improve fracture conductivity.
44
The thorough investigation of conductivity
testing provided a road map of how to
improve breaker effectiveness. The problem
was attacked on two fronts: selection of an
optimal breaker for a given temperature
range and development of an additive to
assist in removal of the degraded polymer.
Each type and composition of breaker are
effective over a certain limited temperature
range, based on performance and cost crite-
ria. This requires tailoring of chemical prop-
erties and encapsulation technology so that
the breaker remains active in the proppant
pack long enough to do its job. A suite of
materials is needed to cover the range of
temperatures that are encountered in frac-
turing operations.
But a breaker by itself is often not enough.
Conductivity testing shows that residual
fragments left after the primary polymer
structure has been degraded can still cause
significant pore blockage (below ) . Worse,
under certain conditions, the fragments can
coagulate—or bind together into a viscous
mass—and reduce conductivity further.
Sometimes, adding more breaker can aggra-
vate coagulation.
300 To avoid this, a special blend of anticoag-
ulants was developed. This blend, the
CleanFLOW additive, works synergistically
with the breaker to reduce the size of the
polymer fragments and prevent their ten-
dency to coagulate. The dispersive action of
the additive increases mobility and available
+ breaker
+ CleanFLOW
and breaker
35
Test 1
Test 2
30
Retained permeability, %
25
20
15
10
5
0
Lab 1 Lab 2 Lab 3
nInterlaboratory comparison of conduc-
tivity results. Major improvements in
reproducibility stem from consistency in
testing methods and close control over
test conditions, removing a major road-
block in correlating data between labs.
flow paths. Pore blockage is reduced and
proppant-pack permeability increases.
Laboratory testing in standard conductiv-
ity cells shows that the breaker plus addi-
tive system outperforms breaker alone,
yielding almost 40% higher proppant-pack
permeabilities at low-end additive concen-
trations (next page, top). Retained perme-
abilities improve to as much as 90% at
increased concentrations.
In field testing, treatments with the
CleanFLOW additive showed higher levels
of polymer returns during flowback than
previous treatments on offset wells. Higher
returns are a direct measure of improved
proppant-pack cleanup. Over a given flow-
back period in a well in western Wyoming,
USA, 62% of the polymer was returned
when the additive was used with breaker,
compared to 31% for an offset well with the
breaker alone. At the end of the test phase,
the well treated with the CleanFLOW addi-
tive was still returning polymer, while the
offset well was not (next page, bottom ).
Net present value (NPV) calculations on
various formation types show that conduc-
tivity-related production increases realized
from the use of the additive can significantly
improve well economics.
Oilfield Review
Reducing Polymer Concentration vated temperatures, however, higher viscosi- els can be reduced to about 25 lbm/1000
Today, over 70% of the fracturing treatments ties have historically been used. In non- gal [3 g/cm3]. The goal of recent research
conducted use guar- or hydroxypropyl guar- crosslinked (linear) systems, this means has been to formulate a reliable fluid using
based fluids.18 The rheology of such fluids adding more polymer, resulting in polymer even less polymer to reduce cost and
has been studied for years. When added to concentrations of 40 lbm/1000 gal [4.8 improve fracture conductivity.
water, guar molecules hydrate and swell, g/cm3] or higher. This approach is expensive Various metal ions, including titanium
increasing in diameter and length. The and often results in fluid mixing and han- and zirconium, have been used for decades
hydrated strands overlap and hinder motion, dling difficulties. Fracture conductivity can as crosslinking agents. In recent years,
giving rise to an increased viscosity of the be impaired, since more polymer must be boron [as B(OH)4-] has grown in popularity
solution. For adequate proppant transport broken and produced back. Crosslinking and is by far the most common element
and placement, a viscosity of 100 centipoise has become a common means of enhancing today. Various boron salts and compounds
(cp) at a shear rate of 100 sec-1 is generally viscosity at lower polymer levels. While
accepted as a minimum guideline. effective in building viscosity, this practice 18. Gulbis J: “Fracturing Fluid Chemistry,” in Econo-
mides MJ and Nolte KG (eds): Reservoir Stimulation,
To minimize fluid leakoff and counteract can lead to complicated, hard-to-break 2nd ed. Englewood Cliffs, New Jersey, USA: Prentice
the inherent thinning of guar systems at ele- structures. Using crosslinkers, polymer lev- Hall (1989): 4-1–4-14.

40 nConductivity
comparison. The
Temperature = 225°F
new additive sys-
tem provides sub-
30 stantially greater
Retained permeability, %

retained perme-
abilities compared
with no breaker or
20 with use of a
breaker alone, con-
firming that the
additive enhances
10
removal of residual
polymer and limits
pore blockages.

0
No breaker Breaker only CleanFLOW CleanFLOW
system system
2 gal/1000 gal 4 gal/1000 gal

40

CleanFLOW system and breaker

nPolymer flow-
35 back comparisons.
Breaker only For a test well in
Polymer concentration, lbm/1000 gal

western Wyoming,
30
USA, 62% of the
polymer was
25 returned during
62% polymer return flowback. For an
20 offset well with
breaker alone, dur-
15 ing a similar flow-
back time, only
10
31% of the polymer
was returned.
31% polymer return
5

0
200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
Cumulative flowback volume, bbl

Autumn 1995 45
can be used (below ) . If the pH19 of the fluid ing extended pumping times. For these 1
75°F
is above 8, crosslinking occurs almost applications, it is necessary to delay the

Borate ion fraction

0.8
100°F
instantaneously when borate ions are added crosslinking process until the fluid has trav-
0.6 150°F
to hydrated guar (right ) . A unique feature of eled through most, or all, of the well tubu-
lars. This can be accomplished by varying 200°F
the resulting fluid is that the crosslinking is 0.4
reversible. As temperature increases, the pH the chemistry and additives in the fluid or by 250°F
falls and the solution thins because there is encapsulating the crosslinker, which permits 0.2 300°F
less borate ion available in solution. Viscos- timed release of the active material. Delayed 0
ity recovers as temperature decreases and crosslinking reduces friction pressure and 5 6 7 8 9 10 11 12 13 14

the pH rises. horsepower requirements and provides for pH

If crosslinking is too rapid, high friction
pressure in wellbore tubulars and shear
thinning may occur during pumping. This
may not be a severe problem in shallow,
low-temperature wells where pumping
times are short, but is of major concern for
deep, higher temperature treatments requir-
higher injection rates.
For a stable and reliable borate crosslinked
fluid to be prepared at a given temperature,
sufficient polymer chains must be present for
entanglement to occur. For most guars, solu-
tions containing less than 20 lbm/1000 gal
cannot effectively be crosslinked. In addition
nBorate ion fraction as a function of tem-
perature and pH. The amount of borate ion
in solution directly impacts the crosslinking
process. As the pH increases past 8, the
amount of borate available for crosslinking
increases dramatically. The solubility
effect gives rise to the reversible crosslink-
ing exhibited by borate fluid systems.

Structure of Guar to having sufficient polymer in solution, two

Galactose
nStructure of guar other criteria must be satisfied: the chains
CH2OH CH2OH and crosslinking need to have enough active crosslink sites
substituents
O O by boron. Guar is and the proper number of borate ions must
HO H HO H composed of a
H H be present to build a network structure.
mannose back-
Careful balance between the two has to be
H OH H
H
OH H
Mannose bone with intermit-
backbone tent galactose sub- maintained to produce a stable system.
O O stituents (top). In the past, borate crosslinked fluids at the
H OH H OH
Boron, in the form lower threshold of 20 lbm/1000 gal guar
CH2 CH2 CH2OH CH2 of B(OH)4-, reacts concentration have been formulated in the
O O O O with the hydroxyl
H H H H H (OH) groups on the laboratory and even tested in the field.
H H
O O O O polymer in a two- Operational experience, however, showed
OH OH H OH OH H OH OH H OH OH H step process to link them to be unreliable and overly sensitive to
two polymer small variations in fluid chemistry.
strands together
H H H H H H H H
(bottom).
How can the threshold polymer level nec-
essary for crosslinking be reduced success-
Borate Crosslinking of Guar fully? Based on earlier work with zirconate
systems, scientists evaluated what types of
materials associate with guar molecules to
B(OH)3 + OH- B(OH)4- increase their solubility. After studying the
behavior of a variety of materials, researchers
identified a combination of chemicals that
OH O OH
permits stable borate crosslinking at polymer
B(OH)4- + R R B + 2H2O
OH
concentrations as low as 15 lbm/1000 gal
O OH
and reliable field formulations to be mixed in
the 15- to 20 lbm/1000 gal range (next page,
OH O O
bottom ).
B(OH)4- + 2R R B R + 4H2O These systems, referred to as low-guar flu-
OH O O
ids, can be used at fluid temperatures up to
R= Mannose backbone 175°F [80°C]. They exhibit viscosities nor-
mally measured for borate crosslinked fluids
with polymer concentrations 5 to 10
lbm/1000 gal higher.
Proppant suspension properties are excel-
lent, allowing use of proppant concentra-
tions as high as 17 to 20 pounds of proppant
added (ppa). The fluid can be continuously
mixed and then crosslinked. It can be effec-
tively degraded using available breaker sys-

46 Oilfield Review
 tems for rapid, efficient cleanup following Permeability Percent Retained nRetained conduc-
Fluid Temperature (°F) (darcies) Conductivity tivity of low-guar
the treatment. Conductivity tests on proppant systems. By reduc-
packs placed with low-guar fluids show 150 2% KCI 100 244 -
ing the polymer
15 lbm 100 148 61
to 200% higher pack permeabilities than concentration,
20 lbm 100 103 42 cleanup is easier
those obtained with conventional polymer 2% KCI 125 -
216 and the retained
concentrations (right ) . 15 lbm 125 130 60 conductivity of the
Over 500 treatments have been pumped 20 lbm 125 106 49 proppant pack is
since the introduction of the fluid in 1994. 30 lbm 125 63 29 greater than for
Because of the success of these treatments, 2% KCI 175 125 - conventional fluids
the low-guar system is rapidly becoming the 20 lbm 175 61 49 that require higher
30 lbm 175 32 26 polymer concen-
fluid of choice for applications to 175°F. trations.
Reduced polymer loading can lead to Proppant Type: Sand
increased fluid leakoff. This can be averted Proppant Concentration: 2 lbm/gal
by the adding fluid-loss agents, particularly
in higher-permeability formations. Field stud- Controlling Proppant Flowback
ies, however, have shown that this may not Flow of proppant into the wellbore follow- widths and higher proppant concentrations
always be necessary, depending on the type ing a fracturing treatment is of major con- have become the norm. In areas such as
of formation and its permeability. In more cern. This phenomenon may occur during Alaska and the North Sea, up to 20% of the
than 130 jobs performed in Kansas, USA in initial cleanup or sometime after the well is proppant may be produced back, while
5- to 50-md formations, no fluid-loss addi- put back on full production. Termed prop- some instances of up to 50% have been
tives were used. Fluid efficiency was not pant flowback or proppant backproduction, reported.20 This can translate into anywhere
adversely affected, and conductive fractures it can lead to expensive, time-consuming from 1000 to 100,000 lbm [454 to 45,400
were obtained. Cleanup times were reduced remedial operations and safety concerns. In kg] of proppant per treatment. Although the
by 50% compared to previous treatments. low-rate wells, proppant may settle in the flowback may stop with time, many wells
Proppant placement efficiencies—the per- casing, requiring periodic wellbore produce proppant throughout their life-
centage of proppant effectively placed in the cleanouts. Loss of near-wellbore fracture times. Wells often must be placed on
fracture—met or exceeded 97%. conductivity can result, and production may restricted chokes to limit pressure drops
In even higher-permeability formations, cease entirely if the productive zone is fully and stabilize the proppant pack.21
as tests in Alberta, Canada (page 42 ) con- covered. In high-rate wells, erosion occurs While changes in fracture design and exe-
clusively demonstrate, low-guar systems to tubulars, control valves and wellhead cution can sometimes alleviate the problem,
can be used effectively with the new equipment. Disposal costs for produced a typical solution has been the use of resin-
HIGHSHEAR fluid-loss additive to improve proppant may be substantial. coated proppant (RCP). RCPs are pumped
fracture conductivity, well cleanup and well The frequency of this problem has into the fracture near the end of the treat-
performance. increased markedly as greater fracture ment, referred to as tailing in. The well may

15 lbm/1000 gal 20 lbm/1000 gal 19. The measure of acid intensity equal to the logarithm
180 400
nViscosity of low- of the reciprocal of the hydrogen ion concentration
of the solution. pH 7 is neutral; below 7 is acidic,
Viscosity at 100 sec-1, cp
Viscosity at 100 sec-1, cp

160 350 guar fluids over above alkaline.

140 300 time. Now, reliable 20. Martins JP, Abel JC, Dyke CG, Michel CM and Stew-
120 fluids can be for- art G: “Deviated Well Fracturing and Proppant Pro-
250 duction Control in the Prudhoe Bay Field,” paper
100 mulated with poly-
200 mer concentrations SPE 24858, presented at the 67th SPE Annual Tech-
80 nical Conference and Exhibition, Washington, DC,
150 of between 15 USA, October 4-7, 1992.
60 100°F
100
lbm/1000 gal (left) 21. Vreeburg R-J, Roodhart LP, Davies DR and Penny
40 125°F and 20 lbm/1000 GS: “Proppant Backproduction During
20 150°F 50 gal (right). Even at Hydraulic Fracturing—A New Failure Mechanism
0 0 the lower polymer for Resin-Coated Proppants,” Journal of Petroleum
0 1 2 3 0 1 2 3 concentration, vis- Technology 46 (October 1994): 884-889.
Time, hr Time, hr cosities are above
the 100-cp level
required for ade-
quate proppant
suspension and
transport.

Autumn 1995 47

be shut in for a period of time to allow the
resin to cure as its temperature rises, binding
the proppant particles together at their
points of contact. Ideally, a consolidated,
high-conductivity matrix is formed.
RCPs, however, are not universally appli-
cable and have certain severe limitations.
Performance is sensitive to shear, tempera-
ture, closure pressure and shut-in time. Con-
ductivities are frequently lower than
expected. In low-temperature wells, an
expensive activator must be added to the
RCP at a 0.5 to 2.0% concentration for the
resin to cure. Shut-in times may be as long
as 24 hours. Resin coatings can interact
chemically with fracturing fluid additives.
Cyclic loadings, caused by the well being
shut in and put back on production over
time, can cause the pack to fail. In extreme
cases, gelled masses of resins can be pro-
duced back.22
Because of these drawbacks, there is a
major incentive to introduce a more consis-
tent means of controlling proppant flowback
that also improves well cleanup efficiency
and maximizes well productivity.
Recent research has helped define the
mechanisms underlying proppant-pack
degradation and has led to the invention of
a physical, rather than a chemical, solution
to the problem. This innovation, called
PropNET technology, uses fibers to hold the
proppant in place. Pumped together with
the proppant in the fracturing fluid, the
48
nPropNET proppant
flowback control.
Random glass
fibers pumped in
the fracturing fluid
form a net-like
structure that stabi-
lizes the proppant
pack while allowing
high production
rates of formation
hydrocarbons.
material forms a web, or network, which
stabilizes the proppant-fiber pack and
allows high production rates of oil or gas
(above ) . The technology is based on the
principles of fiber reinforcement commonly
used in a variety of industrial and commer-
cial applications as a strengthening method.
For example, natural and synthetic fibers are
used to protect dams and other concrete
and soil structures from erosion. The inher-
ent ability of fibers to stabilize highly
porous, particulate-containing materials
provided a basis for these investigations.
A comprehensive set of tests has been
applied to determine the applicability of
fiber reinforcement and to characterize the
properties and performance of fiber-contain-
ing packs. Conductivity test cells measured
proppant-pack permeability. Special pack-
mobility tests simulated conditions before,
during and after well cleanup. Three config-
urations of test cells were developed and
constructed to evaluate the key parame-
ter—pack resistance to proppant flow-
back—as measured by the maximum flow
rate that the pack can withstand or the max-
imum pressure drop across the pack before
proppant is produced.
A variety of fiber types were investigated,
including polymer, glass, ceramic, metal
and carbon. Based on several evaluation cri-
teria, a special, flexible glass fiber was cho-
sen for its performance, cost and availability.
The material has a 2.55 g/cm 3 [21.3
lbm/gal] bulk density, similar to that of most
proppants. Selection of the particular glass
fiber was based on its stability to fluids and
downhole conditions. Fibers must meet
minimum size criteria to be effective. Short,
small diameter fibers are less effective in
strengthening the proppant pack. Glass fiber
diameters of 10 to 20 microns and lengths
of 10 mm or more provide optimum pack
stability and ease of handling.
How the fibers work is open to debate,
but it is thought that they interweave among
the proppant particles, providing increased
strength, or that they stabilize and distribute
stress, aiding bridging, within a significant
area of the pack. The fiber structure is more
flexible than cured RCPs, allowing the prop-
pant-fiber pack to shift without failing.
Laboratory tests show that the ability of
the pack to resist proppant flowback is a
function of fiber concentration. Stability
increases with fiber content until a plateau
is reached (below ). While laboratory data
show that use of 1.5% fibers by weight can
reduce permeability by up to 30% com-
pared to packs without fibers,23 field results
show less reduction. Conductivity values for
packs with fibers are superior to those mea-
sured for postcured RCPs.24
Despite the low concentration of 1.5%,
the fiber level is about 30% by number of
particulates, or about one fiber for every two
proppant particles. Fiber length is an order
of magnitude more than proppant diameter,
so, for example with sand proppants, each
fiber touches approximately five particles
(next page, top).
Single-phase flow tests using water and
two-phase tests with water and gas were
9
Critical flow rate, liter/min
8
7
6
5
4
3
2
1
0
0.5 1 1.5 2 2.5
Weight fraction fiber, %
nEffect of fiber concentration on critical
flow rate. As fiber is added, up to about
1.5% by weight, the flow rate necessary to
fail the pack increases. Beyond 1.5%,
additional fibers do not improve pack
strength significantly.
conducted without a confining stress and
with a stress of 1000 psi. Results show that
pack stability increases with pack width to a
certain limit and that packs with more
spherical and uniform ceramic proppant are
generally less stable than irregular grain
sand packs. Packs can withstand pressure
Oilfield Review
 nPhotomicrograph
of fiber-reinforced
proppant pack.
The glass fibers are
about an order of
magnitude longer
than the diameter
of a typical prop-
pant particle,
allowing each fiber
to contact about
five particles. The
fibers stabilize the
pack by inter-
weaving among
particles, provid-
ing increased
strength and sta-
bility. They may
also promote
bridging and stress
distribution within
the pack.

gradients in excess of 100 psi/ft, but typical the type of minerals present in the water (sil-
140

Maximum pressure drop

levels are 45 psi/ft (right ). Worst-case flow ica being the most important). Results show

before failure, psi/ft

120
conditions (80%/20% gas-to-water ratio) that fibers are expected to retain 50% of
were used to establish maximum flow-rate their effectiveness for at least two years at 100

guidelines. Under these conditions, pack
stability reaches a minimum at about 40%
of the maximum pack strength in single-
phase fluid flow. Based on laboratory data,
maximum cleanup rates of 30 bbl/day/per-
foration [4.8 m3/day/perforation] for sand
and 20 bbl/day/perforation [3.2 m3/day/per-
foration] for ceramic proppants are being
used in the field.
Tests results show that no minimum con-
fining stress, shut-in time or reservoir tem-
perature are required for the fibers to be
effective, overcoming some of the severest
limitations of RCPs.
Packs were cycled under stress to simulate
shut-in and production periods, with packs
subjected to alternating stress levels of 1000
psi and 4000 psi. No pack failures were
observed, even after more than 30 cycles,
for both sand and ceramic proppant packs.
Aging studies were also conducted, since
glass fibers can be dissolved by formation
waters. The solubility rate depends on sev-
eral factors, including temperature, pH and
300°F [149°C] when in contact with a sil- 80
ica-saturated formation brine. Life 60
expectancy could be greater, depending on 40
downhole conditions.
20
Glass fibers do not interact with common
0
fracturing fluid systems or additives, a key
0 20 40 60 80 100
concern with RCPs. Their presence in the Nitrogen content, % total flowing
fluid slurry also reduces proppant settling,
aiding proppant transport and placement. nResistance of sand-fiber packs to prop-
Glass fibers have certain limitations to be pant flowback. At a fiber content of 1.5%,
the pressure drop the pack can withstand
considered during treatment design. They is a maximum when no nitrogen is pre-
are not effective at temperatures above sent, exceeding 100 psi/ft. As the nitro-
300°F and under certain conditions: if for- gen content increases, pack strength
decreases until a minimum is reached at
about an 80% gas volume.
22. Almond SW, Penny GS and Conway MW: “Factors
Affecting Proppant Flowback with Resin Coated
Proppants,” paper SPE 30096, presented at the Euro-
pean Formation Damage Conference, The Hague,
The Netherlands, May 15-16, 1995.
23. This observation is consistent with calculations in
which fibers represent 5 to 6% of the pore volume.
24. Card RJ, Howard PR and Feraud J-P: “A Novel Tech-
nology to Control Proppant Backproduction,” paper
SPE 31007, SPE Production & Facilities (in press).

Autumn 1995 49
 Economic Benefits of Fiber Reinforcement: Field Results

More than 150 hydraulic fracturing treatments using PropNET fiber-reinforced packs have been conducted in
the USA (top) and Venezuela. Here is a look at three typical examples.

Texas, USA:
A 10,000-ft [3080-m] gas well in south Texas,
USA with a bottomhole temperature of 275°F
[135°C] was fractured using a borate-crosslinked
fluid. A fiber-reinforced pack was placed with
1.5% fiber by weight of proppant. Fibers were
added during the entire proppant stage. Pumping
pressure levels were similar to treatments with-
out fibers present. Of the nearly 16,000 lbm
[7300 kg] of proppant placed, less than 0.07%
■Fracture treatment in south Texas.
was produced back. The well was later cycled
through four shut-in and production periods, with
a shut-in closure stress of 1900 psi and a flowing ■ Fluid returns in a
400 PropNET south Texas well.
tubing pressure of 4200 psi. Production was
Faster flowback rates
essentially proppant free and remains so today and larger cumulative
Cumulative fluid returned, bbl
well over a year later. Productivity from the well flowback volumes
300
exceeded that from offsets. were found with the
PropNET system than
In another south Texas gas well, having two Gas
with RCPs. Gas break-
sandstone layers separated by a shale layer, through was much
200
226,000 lbm [102,500 kg] of proppant were Gas
RCP earlier and product
placed using a borate-crosslinked guar fluid. sales were realized
sooner.
Here, fibers were tailed in with the last 15% of 100
the proppant. Flowback started as soon as the
treatment was completed. The initial flowback
rate of 500 bbl/day [80 m3/day] was increased
0
10 20 30 40 50 60 70
to 1000 bbl/day [160 m3/day]. Less than 0.05%
Flowback time, hr
of the proppant was produced back over a
four-day period.
30,000
How fiber reinforcement compares with RCPs ■ Decrease in flowback
was evaluated in an offset well. An upper produc- costs for a south Texas
well. With a much
tive zone was fractured with a 15% fiber tail-in
shortened flowback
Flowback costs, $

treatment, while a lower zone used a 23% RCP 20,000

time, well monitoring
tail-in. The PropNET zone had a much higher ini- Flowback costs
and associated costs
RCP flowback
tial flowback rate, earlier gas breakthrough and decreased by are substantially
$12,600 reduced, providing an
more rapid fluid returns (middle). Cleanup costs 10,000
increased return on
were reduced and fracturing fluid recovery was investment for the
maximized, allowing the well to be placed on PropNET flowback treatment.
production sooner (bottom). Conductivities were
0
2 4 6 8 10 12 14 16
higher with PropNET treatments than for RCPs.
Flowback time, days

50 Oilfield Review

According to Gary Slusher, project production
engineer for Enron Oil & Gas, in Corpus Christi,
Texas, “We have had to live with RCPs. They can
be effective, but they are expensive and require
long cure times, and once you start pumping
them you’re committed. When we bring a well
back, we have to do it gingerly to avoid prob-
lems. With PropNET, we can start and stop pump-
ing as necessary and aggressively flow back the
wells. The system is more effective in stopping
proppant production than RCPs.
“We’ve seen both direct and indirect benefits
with PropNET. We’re getting much higher fluid
recoveries and more rapid cleanup. This reduces
our total job costs substantially, by 12 to 15% or
more. We believe our effective frac lengths are
longer, too. But, most importantly, we get the gas
to market sooner by turning the wells around
faster, sometimes in about 24 hours instead of
seven to ten days. We’ve also been able to use
simpler fluids with less additives. That’s saving
us additional money and reducing friction pres-
sures and horsepower requirements. When the
job has been executed according to design and
the material has been placed where it’s needed,
the results have been excellent.”
Indiana, USA:
In low-temperature gas wells in Indiana, USA,
RCPs are commonly used when fracturing multi-
ple zones. This requires an activator and an
extended shut-in period for curing of each zone
after treatment, or a total time of four to five days
per well. When fiber technology was used
instead, the initial zone could be flowed back
only 10 min after the job was completed. Follow-
ing a limited flowback period, the next zone could
be fractured immediately, completing operations
in one day, a 75 to 80% savings in rig labor, rig
time and equipment costs.
Autumn 1995
mation water is not silica-saturated, if con-
fining stress exceeds the crush strength of
the proppant, or if the fracture will be
treated later with certain types of acid.25
While fibers could be distributed through-
out the entire length of the proppant pack,
field experience supports laboratory tests
showing that only the tail-in portion of the
treatment, typically the last 15 to 20% of the
proppant pumped, normally requires fiber
addition. If, however, there are concerns
that the entire height of the producing zone
cannot be covered or that sand control will
be a problem, it is prudent to use fibers
throughout the proppant pack.
Extensive field experience in the USA
(see “Economic Benefits of Fiber Reinforce-
ment: Field Results,” previous page ) has
demonstrated the superior performance of
fiber-reinforced packs, including reduced
shut-in times, faster cleanup rates and
greater execution efficiency. This translates
directly into rapid well turnaround and
decreased wellsite costs.
On average, less than 0.2% of the prop-
pant pumped was produced back in over
150 treatments. Cleanup rates as high 4000
bbl/day [636 m3/day] water and gas rates of
up to 10 MMscf/day were observed, 10
times those used by local operators for
RCPs. Since cleanup can represent 10 to
60% of the total cost of a fracturing treat-
ment, this time savings has a dramatic eco-
nomic impact.
With the concept proven both in the labo-
ratory and field, this technology will rapidly
be applied beyond the Western Hemi-
sphere. To support this expansion, alterna-
tive fibers are being tested for application
above 300°F in areas such as the North Sea.
New Solutions with Novel Technology
The impressive developments in fracturing
fluids technology from 1985 to 1993 have
been reinforced by innovations during the
past two years. Advanced fluid-loss addi-
tives are improving fluid efficiency at lower
fluid viscosities. Combined with new low-
guar systems, this means reduced costs and
reduced damage to the fracture. Enhanced
breaker and additive systems are speeding
well cleanup, giving more complete fluid
degradation and cleaner, more conductive
fractures. And, the fracture is now more sta-
ble, thanks to innovations in proppant flow-
back control.
Individually and collectively, these new
technologies are benefiting oil and gas oper-
ators by reducing treatment and well
cleanup costs, increasing well productivity
and speeding product sales to market. In the
future, the increased emphasis on more effi-
cient and effective fluids and the synergistic
application of low-cost innovations will
yield further economic gains. —DEO
25. Fibers are resistant to hydrochloric acid [HCl], the
acid commonly used, but not to hydrofluoric acid
[HF] which is known to dissolve glass.
51